Nvidia NIM Reranker
NVIDIA NIM Reranker Utility for Milvus Integration Rerank chunks instead of papers following traditional RAG pipeline
rerank_chunks(chunks, query, config, top_k=25)
Rerank chunks by relevance to the query using NVIDIA's reranker.
This follows the traditional RAG pipeline: first retrieve chunks, then rerank them.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
chunks
|
List[Document]
|
List of chunks to rerank |
required |
query
|
str
|
The query string |
required |
config
|
Any
|
Configuration containing reranker settings |
required |
top_k
|
int
|
Number of top chunks to return after reranking |
25
|
Returns:
Type | Description |
---|---|
List[Document]
|
List[Document]: Reranked chunks (top_k most relevant) |
Source code in aiagents4pharma/talk2scholars/tools/pdf/utils/nvidia_nim_reranker.py
20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 |
|